Roulette theories – New theory is need it

[mr.ore]

I believe "gambler's intelligence" can be coded. What is "gambler's intelligence"? It is a function of all knowledge gambler has about game. Each of these knowledges can be represented as an indicator, and next gambler's decision is a function over all these indicators.

next decision = F(bankroll, target, drawdown, table limits, past wins, past spins, wheel layout - past numbers distance, ap info, ...)

decision is in a format [chance to bet on (ec, dozen, two singles, ...); where to bet (red, 1 dozen, (26,0); how many units]

If we find F, we might have a long term winner. Past spins is for gambler's fallacy simulation. Past wins are somewhat similar, but there is also information how much it was, it would be in format [-20;+35;24],[-10;+70;0],... if player won 70 units ten spins back on 0, and 35 units twenty spins back on 24. That could be useful for money management function, but even money management might be based on gambler's fallacy.

Ok, but now we have got a new theory Long term winner can be described as an function F, and if we find that function, we can win for some time. If there are some events on live wheel which are not as random, then good function could abuse them, even without advatage play information, as long as there exists hidden markov model for them. I mean something like if dealer signature is a reason why is distance betwen spun numbers on wheel more stable from time to time, then the method would recover on this, and even waiting/changing tables would make some sense.

I would start with finding F that wins over all single numbers on Wiesbaden 700000 spins file. Test it on 0, 1, ..., 36, that is 37*700000 spins. Just fitting F to win on them, bet every spin at least 1 unit.

Make a theory, decide a goal, reach that goal, and if we reach that goal, see if such a system works on other random spins, which it is not fitted to. Now I am not sure, if it is even possible to design a system to win over so many spins, lol

[mr.ore]

We cannot be winning roulette forever, it will lose sooner or later. But we can look at that this way:

units_won/units_bet = 36/37 in long term.

It does not mean we are going to be losers at any future spin as long as spin is not infinity. We "just" had to risk more to gain less, that's the nature of this game.

We have to use money management (=nice name for complex progression...).

If there would not be bet limits, game would be practically beatable in finite number of spins.

We have to find a way how to get max out of a table with given limits, and minimize their impact.

[winkel]

...  Now can we please stop the debate about something that we will never agree.

Regards

Nobody has proven G.U.T wrong!

And why do you start a topic and donÂÃ,´t want opposite statements?
DonÂÃ,´t start topics if you want only agreements. 

[RBR7]

Nobody has proven G.U.T wrong!

And why do you start a topic and donÂÃ,´t want opposite statements?
DonÂÃ,´t start topics if you want only agreements.  

As I remember, very respectable member on old VLS, Kon-Fu-Sed simulated and proved that your theory doesn't hold water and then you start arguing and changing facts and in the end you included gamblers intelegence, so that "beast" couldnt be tested and result would be wrong.

Anyway here is the link and all result of simulations: link:://vlsroulette.com/testing-zone/kfs'-gut-test/

I don't want agreements, but for your statments you didn't get any facts and even when I said that I don't want argue anymore and I'm sorry if you understand topic in the wrong way, you still don't want to quit, do you? If your GUT works, do the science paper and make source code for software and prove all the world wrong. If you can't do that or you fier that GUT will fail, then you again didn't give any scientific fact that it works and it is all mumbo-jumbo.

Regards

[mr.ore]

Has anyone programmed GUT in such a way that it produced graphs? There are only overall results of millions of spins, of course they are going to be -2.7%. Can GUT maximise chance, that a winning streak occurs? All what a good method needs is periodic occurence of long enough winning streak. The streak should be on average 2.7% shorter than losing one, of course, but periodicity would be a key for long term profits from light positive progression.

[winkel]

Again:

What KonFuSed proofed was that every possible crossing be over 10 milion spins will gain -2,7%. And that has to be that way, otherwise G.U.T couldnÂÃ,´t be proofed winning!

What he didnÂÃ,´t code was the way to play:

If you play in a trot of 50 spins every possible crossing it is very rare, that you will lose "every" crossing. Most you will win more crossings than you will lose.
And if you just go and watch "what is going on" you can avoid losses.

Ask TwoCatSam!

And I always told that the game described at the old VLS is just the basic not the advanced way.

br
winkel

[mr.ore]

I am looking at it right now, now I have only this, bets are not programmed yet...

[winkel]

In a german forum it is coded nearly to the edge of included "gamblers intelligence"
But it hasnÂÃ,´t been possible to run this complex program over some thousands of spins.

But in some tests the code against me, we both got nearly the same bet-decisions.
And if we were different we both did win the trot.

br
winkel

[mr.ore]

Now betting all crossings triggers until hit or end of session.

[mr.ore]

Hmm, I have it probably wrong. Should I during a session use a trigger several times? I changed time to play a trigger from "till end of session" to 32/numbers_count, so that it always make a profit on a win. Might also change to 37 for a full cycle. If I play that, I do not play it again in that session.

[mr.ore]

Final version for now I think. Bet for 32/number_count on trigger, after that, if same trigger happens again, bet again. Trigger is if there is one more or same of numbers with less hits than those in a higher group. Groups: 0 vs 1, 0 vs 2+, 1 vs 2+. If there are more triggers in one spin, always start bet that with highest sum of count numbers. It might happen that several bets run at once, that might be still a problem. If there is already a trigger running, and a new appears, should I stop playing the old one and play that new? Or run them together?